|
Authors
Michael R Zile, MD
Wilson S Colucci, MD
Section Editor
Stephen S Gottlieb, MD
Deputy Editor
Susan B Yeon, MD, JD, FACC
INTRODUCTION — Diastolic heart failure (HF) (also called heart failure with preserved ejection fraction) refers to a clinical syndrome in which patients have symptoms and signs of HF, normal or near normal left ventricular systolic function, and evidence of diastolic dysfunction (eg, abnormal left ventricular filling and elevated filling pressures) (table 1) [1,2]. Among patients with HF, as many as 40 to 60 percent have a normal or near normal left ventricular ejection fraction (LVEF) [1,3-9]. This condition has been labeled diastolic heart failure (DHF) or "heart failure with normal ejection fraction" [10]. In such patients, diastolic dysfunction is the presumed cause of the HF, which can be confirmed by objective measures [1,11,12].
The treatment and prognosis of patients with DHF will be reviewed here. Issues related to etiology, clinical manifestations, diagnosis, and pathophysiology are discussed separately. (See "Clinical manifestations and diagnosis of diastolic heart failure" and "Pathophysiology of diastolic heart failure" and "Cellular mechanisms of diastolic dysfunction".)
IMPAIRED RESPONSE TO STRESS — Patients with DHF have particular difficulty in tolerating certain kinds of hemodynamic stress:
* They tolerate atrial fibrillation (AF) poorly, since the loss of atrial contraction can dramatically reduce left atrial emptying, LV filling, and LV stroke volume. (See "Hemodynamic consequences of atrial fibrillation and cardioversion to sinus rhythm".)
* They do not tolerate tachycardia well, since the increase in heart rate shortens the duration of diastole and truncates the important late phase of diastolic filling.
* Elevations in systemic blood pressure, especially the abrupt, severe, or refractory elevations often seen with renovascular hypertension, increase left ventricular wall stress, which can worsen myocardial relaxation in patients with DHF. (See "Treatment of hypertension in heart failure", section on 'Diastolic dysfunction' and "Who should be screened for renovascular or secondary hypertension?".)
* The acute induction or worsening of diastolic dysfunction by ischemia raises left atrial and therefore pulmonary venous pressure. This explains why many patients with coronary heart disease (CHD) have respiratory symptoms with their anginal pain, including wheezing, an inability to take a deep breath, shortness of breath, and overt pulmonary edema. These respiratory symptoms can occur in the absence of anginal pain and are often referred to as "anginal equivalents."
Episodes of hemodynamic decompensation may result in pulmonary congestion or edema severe enough to be life-threatening. This phenomenon, called flash pulmonary edema, is discussed in detail separately. (See "Evaluation of acute decompensated heart failure".)
TREATMENT — The treatment of HF due to diastolic dysfunction remains empiric, since trial data are limited.
General principles — Guidelines for treatment of patients with DHF were published in 2005 by the ACC/AHA task force on chronic HF (table 2) [13]. It was concluded that the weight of evidence supported only four modalities:
* Control of systolic and diastolic hypertension
* Control of ventricular rate in patients with atrial fibrillation
* Control of pulmonary congestion and peripheral edema with diuretics
* Coronary revascularization in patients with CHD in whom ischemia is judged to have an adverse effect on diastolic function
A clinical practice review concurred with the ACC/AHA guidelines and provided specific examples of clinical therapies appropriate to achieving these goals (table 3A-B) [2].
Choice of medications — The choice of medications in patients with diastolic dysfunction is determined by two factors:
* Treatment of specific underlying processes such as hypertension or symptomatic CHD. Combined therapy may be warranted since hypertrophied hearts are more sensitive to the deleterious effects of ischemia on LV relaxation than nonhypertrophied hearts [14]. (See "Pathophysiology of diastolic heart failure".)
* The possibly beneficial effect of the drug on the pathophysiology of DHF.
The guidelines concluded that efficacy was less well established for the administration of specific drugs, such as beta blockers, angiotensin converting enzyme inhibitors, angiotensin II receptor blockers, and calcium channel blockers [13].
An important caveat is that the patient who has LV diastolic dysfunction with a small, stiff left ventricular chamber is particularly susceptible to excessive preload reduction, which can lead sequentially to underfilling of the LV, a fall in cardiac output, and hypotension. In patients with severe left ventricular hypertrophy (LVH) due to hypertension or hypertrophic cardiomyopathy, excessive preload reduction can also create subaortic outflow obstruction.
For these reasons, the administration of diuretics or venodilators such as nitrates, dihydropyridine calcium channel blockers, and ACE inhibitors must be performed with caution. Careful attention is required for symptoms of ventricular underfilling such as weakness, dizziness, near syncope, and syncope.
Digoxin is generally NOT used in patients with DHF because contractility is intact. The DIG ancillary trial, a parallel study to the DIG trial, evaluated the role of digoxin in patients with HF and an LVEF >45 percent [15,16]. At a mean follow-up of 37 months, digoxin had no effect on all-cause or cause-specific mortality, or all-cause or cardiovascular hospitalization [15]. (See "Use of digoxin in heart failure due to systolic dysfunction".)
Antihypertensive therapy — Lowering the systemic blood pressure was associated with reduced rate of HF in some large randomized hypertension treatment trials, particularly those including diuretic therapy [17,18]. The choice of a specific antihypertensive agent must be individualized in the presence of coexisting diseases such as diabetes mellitus or chronic obstructive pulmonary disease. (See "Choice of therapy in essential hypertension: Recommendations" and "Indications for use of and contraindications to specific antihypertensive drugs".)
Regression of LVH is an important therapeutic goal, since it may improve diastolic function [19]. A meta-analysis published in 2003 attempted to evaluate the relative efficacy of different antihypertensive drugs for their ability to reverse LVH in patients with hypertension [20]. Eighty trials that included 146 and 17 active treatment and placebo arms, respectively, were evaluated. After statistical adjustments for length of therapy and degree of blood pressure lowering, the relative reductions in left ventricular mass index were (graph 1):
* Angiotensin II receptor blockers (ARBs) - 13 percent
* Calcium channel blockers - 11 percent
* ACE inhibitors - 10 percent
* Diuretics - 8 percent
* Beta blockers - 6 percent
ARBs, calcium channel blockers, and ACE inhibitors produced significantly more regression than beta blockers. The clinical significance of this difference is uncertain since there is as yet no evidence that more rapid regression of LVH is associated with improved long-term outcomes.
The choice of antihypertensive therapy is usually based upon other factors. (See "Choice of therapy in essential hypertension: Recommendations".)
Chronic AF — LV filling in diastolic HF occurs primarily in late diastole and is therefore more dependent than normal hearts on atrial contraction. Tachycardia is also deleterious by shortening the time of diastole. (See "Clinical manifestations and diagnosis of diastolic heart failure", section on 'Pathophysiology'.)
For these reasons, restoration and maintenance of sinus rhythm is preferred when AF occurs in patients with diastolic HF. When this cannot be achieved, rate control becomes important. Beta blockers and calcium channel blockers are the usual first-line agents, with digoxin most often being used in patients with systolic HF. (See "Antiarrhythmic drugs to maintain sinus rhythm in patients with atrial fibrillation: Recommendations" and "Control of ventricular rate in atrial fibrillation: Pharmacologic therapy".)
A combination of these drugs may be required to achieve adequate heart rate control. It is important to measure heart rate during moderate exercise and not to base heart rate control solely on values obtained in the resting state.
An important component of the management of atrial fibrillation, regardless of whether rhythm control or rate control is chosen, is anticoagulation to prevent systemic embolization. (See "Rhythm control versus rate control in atrial fibrillation".)
Antiischemic therapy — Myocardial ischemia often precipitates and/or contributes to diastolic HF. Ischemia can result from CHD and/or LVH with subendocardial ischemia. In such patients, beta blockers and calcium channel blockers are preferred. Nitrates are also effective but, as mentioned above, the reduction in preload can lead to hypotension in selected patients. (See "Overview of the management of stable angina pectoris".)
Revascularization by percutaneous coronary intervention or coronary artery bypass graft surgery may be required in patients with drug-resistant ischemic DHF. (See "Medical versus interventional therapy in the management of stable angina pectoris".)
Beta blockers — Beta blockers have a variety of potential beneficial effects in patients with DHF including slowing the heart rate (which increases the time available for both LV filling and coronary flow, particularly during exercise), reducing myocardial oxygen demand, and, by lowering the blood pressure, causing regression of LVH [21,22]. Slowing the heart rate is particularly important in the treatment of pulmonary congestion due to ischemic DHF and, as noted above, in patients in atrial fibrillation.
In the small SWEDIC trial, 113 patients with symptoms of HF, normal systolic LV function, and abnormal diastolic function were randomly assigned to treatment with carvedilol or placebo, with echocardiographic assessment at baseline and six months [23]. Carvedilol therapy resulted in a significant improvement in the E/A ratio, but no significant improvement in deceleration time, isovolumic relaxation time, or pulmonary vein flow velocity.
Beta blockers may also directly improve diastolic function in patients with idiopathic dilated cardiomyopathy and those with an acute myocardial infarction; both are settings in which beta blockers are recommended because they improve survival [24,25]. Slowing of the heart rate may contribute to this effect by providing more time for calcium exit from myocytes, thereby reversing the cellular calcium overload characteristic of diastolic dysfunction. (See "Cellular mechanisms of diastolic dysfunction".)
However, direct evidence of the clinical efficacy of beta blocker therapy in patients with DHF is lacking. While the SENIORS trial suggested that nebivolol (a beta blocker with vasodilating properties) may be beneficial in patients with HF and preserved systolic function, an OPTIMIZE-HF registry study found no clinical benefit of beta blocker therapy in this population.
* In the SENIORS trial, 2128 patients ≥70 years of age with history of HF or known LVEF ≤35 percent were randomly assigned to nebivolol or placebol [26]. Nebivolol therapy resulted in reduction in the primary outcome of all cause mortality or cardiovascular hospital admission (31 versus 35 percent) at mean 21 month follow-up. While the majority of study patients had LVEF ≤35 percent, there was no significant influence of ejection fraction on the effect of nebivolol on the outcome. The effect of nebivolol on outcomes was similar in those with preserved and impaired LVEF [27].
* On the other hand, a difference in the clinical efficacy of beta blocker therapy in patients with systolic HF versus patients with diastolic HF was suggested by a retrospective study from the OPTIMIZE-HF registry of 7154 elderly adults hospitalized with HF [28]. Among patients with left ventricular systolic dysfunction, beta blocker therapy was associated with reduced mortality and rehospitalization rates. In contrast, among patients with preserved systolic function, beta blocker therapy was associated with no improvement in mortality or rehospitalization.
At present, there is no good demonstration that beta blockade is beneficial for the treatment of heart failure with preserved ejection fraction.
Calcium channel blockers — Calcium channel blockers, especially verapamil, also may be useful in the treatment of diastolic HF. These agents have been purported to have a direct "lusitropic" (relaxation-enhancing) effect. However, it is difficult to distinguish this effect from the benefits related to slowing of the heart rate, both at rest and during exercise, and to a reduction in or prevention of ischemic episodes. (See "Calcium channel blockers in the management of stable angina pectoris".)
The following observations from small studies are consistent with the possible efficacy of verapamil:
* In a small randomized, crossover trial of 20 patients with diastolic HF, verapamil, compared to placebo, significantly reduced the signs and symptoms of HF and increased LV diastolic filling rate and treadmill exercise time [29].
* In patients with hypertrophic cardiomyopathy (HCM), verapamil can improve LV diastolic function and may prolong long-term survival [30,31]. In a report of 55 patients with HCM who were treated with verapamil (360 to 480 mg/day) for one to four weeks, an increase in peak LV diastolic filling rate occurred in 43 (78 percent) [30]. Objective symptomatic improvement on graded exercise testing occurred in 34 of these patients compared to only 1 of 12 without an increase in diastolic filling rate (79 versus 12 percent). These benefits persisted for one to two years, and were reversed when therapy was discontinued. (See "Medical therapy in hypertrophic cardiomyopathy".)
ACE inhibitors — There is no clear evidence from randomized clinical studies that ACE inhibitor therapy directly improves overall morbidity or mortality in patients with diastolic heart failure. Because patients with diastolic heart failure frequently have co-morbidities such as renal insufficiency, ACE inhibitors should be used carefully to avoid the risk of renal dysfunction and hypotension. Despite these concerns, ACE-I play an important role in the treatment of the disease processes that underlie the development of diastolic heart failure, namely, hypertension, coronary artery disease and diabetes.
* ACE inhibitors are beneficial in hypertensive heart disease. The reduction in systemic pressure can lead to regression of left ventricular hypertrophy (LVH) and a gradual improvement in diastolic function (see 'Antihypertensive therapy' above [20,32].
* ACE inhibitors are beneficial in patients with mixed systolic and diastolic HF. They are indicated in patients with systolic HF, since they slow the rate of progression of the myocardial disease and improve survival. In such patients, ACE inhibitors also prevent progression of diastolic dysfunction [33]. (See "ACE inhibitors in heart failure due to systolic dysfunction: Therapeutic use".)
The more general efficacy of ACE inhibitors in patients with diastolic heart failure was assessed in the PEP-CHF trial in which 850 patients ≥70 years of age with diastolic dysfunction: 79 percent had a history of hypertension; patients with substantial LV systolic dysfunction or valve disease were excluded [34]. The patients were randomly assigned to an ACE inhibitor (perindopril) or placebo. At one year, treatment with perindopril was associated with an almost significant trend toward reduction in the primary end point of combined all-cause mortality and unexpected hospitalization for HF (8.0 versus 12.4 percent, HR 0.69; 95% CI 0.47-1.01); this effect was entirely due to fewer unexpected hospitalizations for HF. The patients treated with perindopril also had significant improvements in functional class and six minute walk distance.
While benefit from ACE inhibitors may be mediated by its blood pressure effects, it is also possible that there might be a reduction in myocardial angiotensin II production, which can diminish myocardial stiffness. Evidence in support of this hypothesis has been provided by a study of patients with diastolic HF due to aortic stenosis-induced LVH [35]. The intracoronary infusion of enalaprilat, which produced no evidence of systemic ACE inhibition but presumably impaired cardiac angiotensin II production, improved diastolic distensibility and relaxation. Similar benefits may occur in patients with ischemic heart disease [36] and during experimental low-flow ischemia in hypertrophied hearts with ischemic diastolic dysfunction [14]. (See "Actions of angiotensin II on the heart" and "Angiotensin converting enzyme inhibitors and receptor blockers in heart failure: Mechanisms of action".)
Angiotensin II receptor blockers — There is no clear evidence from randomized clinical studies that ARB therapy directly improves overall morbidity or mortality in patients with diastolic heart failure. There is no clear evidence of improved diastolic function with angiotensin II receptor blocker treatment as compared to other therapies in patients with asymptomatic left ventricular diastolic dysfunction or overt diastolic HF.
Angiotensin II receptor blockers (ARBs) may be beneficial in patients with left ventricular diastolic dysfunction due to hypertension, although the data on impact compared to other agents are limited [19,20,37-40]. In various studies of hypertensive patients with LVH, ARBs caused more regression of LVH than beta blockers [20,38], a change that was associated with an improvement in LV diastolic filling [19]. In other studies, losartan treatment improved exercise tolerance and quality of life (which were not improved by hydrochlorothiazide) [39] and reduced myocardial fibrosis and stiffness [40]. (See "Clinical implications and treatment of left ventricular hypertrophy in hypertension", section on 'Effect of antihypertensive therapy'.)
However, a large randomized trial in patients with hypertension and diastolic dysfunction demonstrated that improvement in diastolic function with angiotensin II receptor blocker therapy was similar to that with other antihypertensive therapy [41]. The trial included 382 patients who were treated with antihypertensive drugs not related to angiotensin inhibition and then randomly assigned to valsartan or placebo [41]. There were similar improvements in the two groups in blood pressure control and diastolic function, as measured by myocardial relaxation velocity of the lateral mitral annulus (E').
Outcomes of treatment with angiotensin II receptor blockers in patients with HF and preserved left ventricular ejection fraction (LVEF) were investigated in two large trials:
* In the CHARM-Preserved trial, 3023 patients with symptomatic HF (nearly all NYHA class II or III) and an LVEF >40 percent were randomly assigned to either candesartan (mean dose at six months 25 mg) or placebo [42]. The mean LVEF was 54 percent. Additional medications included ACE inhibitor in 19 percent, beta blocker in 56 percent, and calcium channel blocker in 31 percent.
At a median follow-up of 37 months, there was a small and almost significant difference in incidence of the primary end point of cardiovascular death or hospitalization for HF (22 versus 24 percent; adjusted hazard ratio 0.86; 95% CI 0.74-1.00) that was entirely due to a significant reduction in hospitalization for HF with candesartan (16 versus 18 percent). The overall rate of discontinuation of therapy was similar in the two groups (22 versus 18 percent). However, significantly more patients discontinued candesartan because of renal dysfunction, hyperkalemia, or hypotension.
* In the I-PRESERVE trial, 4128 patients with symptomatic HF (nearly all NYHA class II or III) and an LVEF ≥45 percent were randomly assigned to either daily irbesartan 300 mg or placebo [43]. The mean LVEF was 59 percent. Additional medications included ACE inhibitor in 26 percent, beta blocker in 59 percent, and calcium channel blocker in 40 percent.
At a mean follow-up of 49.5 months, there was no significant difference in the primary end point of death from any cause or hospitalization for a cardiovascular cause. There were also no significant differences in secondary outcomes which included death from HF or hospitalization for heart failure, death from any cause, hospitalization for a cardiovascular cause, and quality of life.
Aldosterone antagonists — Aldosterone contributes to cardiac hypertrophy and fibrosis [44,45]. These processes may be preventable or even reversible by aldosterone blockade [46-48].
The possible benefit of aldosterone antagonism in diastolic HF was suggested in a study of 30 medically treated patients with exertional dyspnea, an LVEF >50 percent, and diastolic dysfunction who were randomly assigned to either spironolactone 25 mg/day or placebo [49]. At six months, spironolactone therapy was associated with significant improvement in myocardial function by echocardiographic indices, including strain rate, peak systolic strain, and cyclic variation of integrated backscatter.
A large NIH sponsored study, TOPCAT, is evaluating the hypothesis that spironolactone is beneficial in patients with a normal ejection fraction and heart failure.
Statins — Intensive lipid lowering with statin therapy is recommended for the secondary prevention of cardiovascular disease, independent of the presence of diastolic dysfunction. (See "Intensity of lipid lowering therapy in secondary prevention of coronary heart disease" and "Treatment of lipids (including hypercholesterolemia) in secondary prevention".)
Initial observational data suggest that statins might be of benefit in patients with diastolic HF.
Randomized trials are required to confirm these observations. Statin therapy in patients with HF is discussed in detail separately. (See "Statin therapy in patients with heart failure".)
Exercise conditioning — During exercise in healthy individuals, diastolic function is enhanced so that left ventricular input remains precisely matched to LV output, despite the shortened duration of diastole resulting from the associated tachycardia. This is achieved in the normal LV by a rapid and marked decrease in intraventricular pressure during early diastole, thereby creating a greater LV "suction" effect, which enhances the transmitral pressure gradient without increasing left atrial pressure and compromising pulmonary function. This mechanism is lost in patients with diastolic dysfunction; as a result, dyspnea with exertion is often their most common complaint. (See "Pathophysiology of diastolic heart failure", section on Left ventricular diastolic function during exercise.)
Long-term exercise training produces physiologic cardiac hypertrophy with enhanced diastolic function. Experimental studies suggest that exercise conditioning has the potential to reverse the diastolic dysfunction of pathologic LVH or aging [50-53]. It is not known if exercise training is beneficial in patients with diastolic HF. However, some observations support such a benefit. In one report comparing 20 distance runners with 20 sedentary individuals matched for age and systolic and diastolic blood pressure, the exercise group had less of the left ventricular diastolic dysfunction associated with "normal" aging [54].
Any exercise training program for the potential treatment of diastolic HF should be based upon dynamic isotonic exercise, not static exercise, since the latter causes changes in cardiac geometry similar to those of hypertensive LVH [55]. (See "Components of cardiac rehabilitation and exercise prescription".)
PROGNOSIS — The prognosis in patients with diastolic HF differs from patients with asymptomatic diastolic dysfunction.
Symptomatic diastolic HF — The prognosis of patients with symptomatic diastolic HF is less well defined than in those with systolic HF. Data from the Framingham Heart Study, the V-HeFT trials, and several other observational series revealed varying results [7,9,56-61]. In addition, the data are not clear on whether the long-term prognosis differs between diastolic and systolic HF [62].
In reports from the community-based Framingham Heart study and the larger Cardiovascular Health Study, prognosis was significantly better in patients with diastolic dysfunction [56,58]. In the latter study of 269 patients at least 65 years of age who had HF, the mortality rates were 87 versus 154 deaths per 1000 person-years with normal and impaired systolic function, respectively (adjusted hazard ratio 1.48 and 1.88 compared to persons with no HF and normal left ventricular systolic function) [58].
In contrast, another community-based study from Olmsted County, Minnesota found that, after adjustment for risk factors, five-year survival was not significantly different in patients with diastolic versus systolic HF (adjusted relative risk 0.80, p = 0.37) [7].
Among patients hospitalized for HF, the mortality rates are higher but the data are again conflicting as to whether or not the prognosis is different in diastolic and systolic HF:
* Among 6076 patients discharged from a Mayo Clinic Hospitals in Olmsted County, Minnesota with a diagnosis of decompensated HF over a 15 year period (1987-2001), 53 percent had a reduced LVEF and 47 percent had a preserved LVEF [9]. One-year mortality was relatively high in both groups, but slightly lower in patients with a preserved LVEF (29 versus 32 percent in patients with reduced LVEF, adjusted hazard ratio 0.96, 95% CI 0.92-1.00). Survival improved over time for those with reduced left ventricular ejection fraction (LVEF) but not for those with preserved LVEF.
* In a prospective evaluation of 413 patients hospitalized for HF, the relative risk for six month mortality was lower for diastolic versus systolic HF dysfunction (13 versus 21 percent, adjusted hazard ratio 0.51) [60].
* In a cohort of 2802 patients discharged from 103 hospitals in Ontario with a diagnosis of decompensated HF, one-year mortality was 22 percent in patients with a preserved LVEF versus 26 percent in patients with a reduced LVEF [61]. This difference was not statistically significant.
Independent predictors of mortality in patients with diastolic heart failure in different studies include older age, male gender, NYHA class, lower LVEF, the extent of coronary artery disease, peripheral arterial disease, diabetes, impaired renal function, the degree of diastolic dysfunction as assessed by Doppler echocardiography, and increased red cell distribution width [63-68].
Asymptomatic diastolic dysfunction — Three studies have assessed the impact of asymptomatic diastolic dysfunction on prognosis:
* In the Mayo Clinic cross-sectional community survey of 2042 adults ≥45 years of age, 21 percent had mild diastolic dysfunction, 7 percent had moderate diastolic dysfunction, and 1 percent had severe diastolic dysfunction [4]. At a median follow-up of 3.5 years, 48 subjects died. After controlling for age, sex, and the presence of systolic dysfunction, all-cause mortality was increased in patients with mild diastolic dysfunction (95 percent of whom were asymptomatic; hazard ratio 8.3) and in those with moderate to severe diastolic dysfunction (90 percent of whom were asymptomatic; hazard ratio 10.2).
* In another report, 3008 American Indians, 45 to 74 years of age, underwent Doppler echocardiography to determine the E/A ratio, and were followed for three years [69]. The 16 percent of patients with an E/A ratio <0.6 (impaired diastolic relaxation) and the 3 percent with an E/A ratio >1.5 (restrictive pattern due to reduced compliance) (graph 2) had a significantly higher all-cause mortality (12 and 13 versus 6 percent for normal E/A ratio) and cardiac mortality (4.5 and 6.5 versus 1.6 percent). After adjustment for covariates, an E/A ratio >1.5, but not an E/A ratio <0.6, was independently associated with all-cause and cardiac mortality (relative risks 1.7 and 2.8, respectively).
* The prevalence and prognosis of asymptomatic diastolic dysfunction was also assessed in 693 patients with coronary heart disease, normal systolic function, and no history of HF who were followed for three years [70]. Diastolic function was normal in 66 percent, mildly impaired in 24 percent, and moderately or severely impaired in 10 percent. After multivariable adjustment, the presence of moderate to severe diastolic dysfunction was strongly predictive of incident hospitalization for HF (hazard ratio [HR] 6.3) and death from heart disease (HR 3.9).
INFORMATION FOR PATIENTS — Educational materials on this topic are available for patients. (See "Patient information: Heart failure causes, symptoms, and diagnosis" and "Patient information: Heart failure treatments".) We encourage you to print or e-mail these topic reviews, or to refer patients to our public web site, www.uptodate.com/patients, which includes these and other topics. |
